Journal of Pediatric Psychology, Vol. 27, No. 7, 2002, pp. 607-617
© 2002 Society of Pediatric Psychology
Predictors of PTSD in Mothers of Children Undergoing Bone Marrow Transplantation: The Role of Cognitive and Social Processes
1 Fox Chase Cancer Center, 2 Ruttenberg Cancer Center, Mt. Sinai School of Medicine, 3 Memorial Sloan-Kettering Cancer Center, 4 Dana Farber Cancer Institute, 5 Children's Memorial Hospital, Northwestern University Medical Center, 6 Packard Children's Hospital, Stanford University Medical Center, 7 Emory University Medical Center
All correspondence should be sent to Sharon Manne, Division of Population Science, Fox Chase Cancer Center, 7701 Burholme Avenue, CPP Suite 1100, Philadelphia, Pennsylvania 19111. E-mail: SL_Manne{at}fccc.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Objective: To investigate the role of cognitive and social processing in posttraumatic stress symptoms and disorder (PTSD) among mothers of children undergoing bone marrow and hematopoietic stem-cell transplantation (BMT/SCT).
Method: Questionnaires assessing emotional distress, BMT-related fears, and negative responses of family and friends were completed by 90 mothers at the time of the BMT infusion and 3 and 6 months post-BMT. PTSD symptoms were measured 6 months post-BMT by both paper-and-pencil and structured interview methods.
Results: Emotional distress, BMT-related fears, and negative responses of family and friends assessed at the time of BMT hospitalization were predictive of later PTSD symptoms. None of these variables prospectively predicted a PTSD diagnosis as measured by the structured interview.
Conclusions: Higher levels of general psychological distress, cognitive interpretations of the threat of the BMT for the child's future functioning, and negative responses of family and friends may place mothers at risk for post-BMT posttraumatic stress symptomatology.
Key words: posttraumatic stress disorder; pediatric bone marrow transplantation; mothers; psychological distress.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Bone marrow transplantation (BMT) and hematopoietic stem-cell transplantation (SCT) have become standard therapy for many childhood diseases. Unfortunately, these procedures are risky, and mortality rates are as high as 50% a year after BMT (Balduzzi et al., 1994
). Even if treatment is successful, there is a risk of
recurrence, chronic graft-vs.-host disease, and serious late effects (e.g.,
pulmonary disease and cardiac problems) (Leisner, Leiper, Hann, &
Chessells, 1994). BMT/SCT can also have a profound psychological impact on
parents. Parents may experience anxiety and depressive symptoms
(Barrera, Boyd, Pringle, Sumbler, &
Saunders, 2000), as well as post-traumatic stress disorder (PTSD).
PTSD symptoms include intrusive thoughts or reexperiencing aspects of the
traumatic event, avoidance of reminders of the event or a numbing of emotions,
and hypervigilance or increased physiological arousal. Studies evaluating PTSD
in parents have reported incidence rates between 6% and 39.5%
(Butler, Rizzi, & Handwerger,
1996; Kazak et al.,
1997; Manne, DuHamel, Galleli, Sorgen, & Redd, 1998;
Pelcovitz et al., 1996).
However, few studies have evaluated PTSD among parents of children undergoing
BMT. Heiney, Neuberg, Myers, and Bergman
(1994) studied 20 parents of
children undergoing BMT and found that parents described intrusive memories
about the BMT and child medical symptoms that prompted fear and anxiety.
Why some parents experience PTSD and others do not is not well understood.
Theory and research suggest that cognitive and social processing are involved
in PTSD (Lang, 1985;
Tait & Silver, 1992).
Cognitive processing refers to how the individual comes to understand the
implications of the event, in both cognitive and emotional terms, and social
processing refers to the way that the experience is talked about and dealt
with in the individual's social network.
Cognitive processing theories suggest factors that place some parents at
greater risk for PTSD. Creamer and others
(Creamer, Burgess, & Pattison,
1992; Lang, 1985)
have proposed that a traumatic event results in the formation of a fear
network composed of cognitive, affective, and physiological responses and
interpretive information about the event (e.g., its meaning). Network size is
determined by the perceived threat to the person's life, the intensity of the
person's fear reaction, and the appraised potential for personal suffering
invoked by the event. The first dimension of the fear network, perceived
threat, has been associated with PTSD in studies of individuals who had
observed a multiple shooting (Creamer et
al., 1992) and studies of parents of pediatric cancer survivors
(Kazak et al., 1998). The
second dimension of the fear network, the intensity of the person's fear
reaction, has also been found to predict PTSD. Immediate emotional reactions
have been most frequently studied through analysis of retrospective ratings of
the level of fear (Creamer et al.,
1992) or anxiety and horror during the event
(Cardena, Koopman, Classen, Waelde, &
Spiegel, 2000; Maercker,
Beauducel, & Schutzwohl, 2000). Unfortunately, research has
not addressed the role of the third component of the fear network, the
potential for personal suffering, in the development of PTSD.
A second relevant theory is social processing, which postulates that
talking with others can facilitate cognitive and affective processing of
traumatic life experiences (e.g., Lepore,
Silver, Wortman, & Wayment, 1996;
Tait & Silver, 1992). As
Tait and Silver suggest, talking with others may facilitate recovery through
discharging of emotions, learning to tolerate aversive feelings, and receiving
support and encouragement of effective coping. Not being able to talk, either
because others act uncomfortable or are overtly critical, may place
individuals at higher risk for PTSD, either because they cannot derive the
benefits outlined above, or because these responses lead to avoidance. Our
cross-sectional research among mothers of childhood cancer survivors has
suggested that supportive responses are associated with lower levels of PTSD
symptoms, while negative responses are associated with higher levels of
symptoms (Manne et al.,
2000).
Finally, objective medical aspects of transplantation may play a role in
PTSD because they represent greater exposure to aversive aspects of the
trauma, which increase perceived potential harm (e.g.,
March, 1993). Parents of
children undergoing BMT/SCT who experience aversive events such as a
life-threatening infection may also be more likely to develop PTSD. Although
this hypothesis has not been supported in studies of parents of pediatric
cancer survivors (Kazak et al.,
1998), it should be evaluated because it has been associated with
PTSD in other studies (e.g., Kliewer,
Lepore, Oskin, & Johnson, 1998).
In this longitudinal, prospective study, three sets of factors were
investigated: cognitive processing (e.g., fear network), social processing,
and BMT-related medical events and treatment severity factors, in PTSD
symptoms and diagnosis. Five study predictions were made. First, we predicted
that mothers who had a higher fear network appraisal (perceived threat to the
child's life, potential for the child's suffering, and fears for the child's
future) during the BMT and at 3 months post-BMT would be more likely to report
PTSD 6 months after the BMT. Second, we predicted that mothers experiencing
more general distress during and 3 months post-BMT would be more likely to
report PTSD. Third, we predicted that unsupportive responses from others and
lower levels of social support would be associated with PTSD symptoms. Fourth,
we predicted that mothers who were exposed to severe medical events or had
children who underwent a poorer prognosis BMT would be more likely to report
PTSD. We focused on mothers who, because of their role as primary caretaker,
may carry greater risk for distress reactions during their child's BMT/SCT
(Streisand, Rodrigue, Houck, Graham-Pole,
& Berlant, 2000).
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Participants
Participants were 90 mothers of children undergoing BMT or SCT. This investigation uses data from participants in a multisite longitudinal study of maternal distress and coping after pediatric BMT. Potential participants were recruited from BMT units at six hospitals (Mount Sinai School of Medicine, Memorial Sloan-Kettering Cancer Center [MSKCC], Dana Farber Cancer Institute, Northwestern University Medical Center, Stanford University Medical Center, and Emory University Medical Center). Inclusion criteria were (1) mother can read and write English, (2) mother was the primary caretaker, (3) mother age 18 years or greater, (4) child age 21 years or younger, and (5) mother had a partner. We approached 226 eligible mothers. Seventy-eight mothers (34%) declined participation and 143 (63%) consented. The most common reasons for nonparticipation were being overwhelmed (n = 15), lack of interest (n = 13), not wanting to leave the child's bedside (n = 11), and study burden (n = 7). Six of the consenting mothers (4.1%) were unable or unwilling to complete the Time 1 assessment and therefore could not be included. One hundred forty-three mothers completed the Time 1 assessment. Of these, 22 did not complete the Time 2 or 3 assessments (15.3%). Thirty-one children died prior to Time 3 (21.6%). The final sample consisted of 90 mothers (63% of mothers completing Time 1).
The mean age of participants was 37 years, 6 months (SD = 7.35, range = 19-56). Most were Caucasian (75.6%). Seventy-seven (85.6%) were married. Median income level was $50,000 to $59,999, and 84.4% had had at least some college education. Children ranged in age from 9 months to 20 years, with a mean age of 8.75 years (SD = 5.51). Approximately half were males (57.8%). Time since the child's diagnosis ranged from 1.5 months to 10 years, 9 months, with a median of 7 months. The most frequent cancer diagnoses were acute leukemias (n = 39) and neuroblastoma (n = 15). The most frequent noncancer diagnoses were sickle cell disease (n = 3) and juvenile rheumatoid arthritis (n = 1). Sixty-six children (73.3%) were undergoing an allogeneic BMT, and 24 (26.7%) were autologous BMT. Children were hospitalized an average of 27 days posttransplant (range = 7-88 days).
Analyses comparing participants and study refusers were conducted.
Demographic categories were collapsed to create the following dichotomous
variables: married/nonmarried, Caucasian/non-Caucasian, cancer
diagnosis/noncancer diagnosis. In addition, comparisons of hospital site,
child gender, type of transplant (autologous/allogeneic), and child age were
conducted. Results did not suggest differences between groups with regard to
child gender, cancer versus noncancerous condition, ethnicity, or marital
status. There were significant differences in the percentage of refusers
across study sites (
2 [n = 242, df = 5] =
12.4, p < .05). The three study sites that had larger BMT programs
had a higher percentage of refusers (Dana Farber, 33%; MSKCC, 32%;
Northwestern, 38%) than the smaller study sites (Stanford, 9%; Emory,
12%).
Analyses were also conducted to compare participants with the 22 mothers who did not complete Time 3 (for reasons other than the child's death) on Time 1 variables. Mothers who did not complete Time 3 had more anxiety (t [104] = 3.5, p < .01), more depressive symptoms (t [104] = 2.5, p < .05), and greater "fear network" (t [131] = 3.1, p < .01).
Procedures
Eligible mothers were approached prior to the child's BMT/SCT. After
informed consent was obtained, mothers completed paper-and-pencil
questionnaires at three time points and one structured interview at the third
time point. The first assessment (Time 1) took place after the child was
admitted to the hospital for the BMT. The mean date of the Time 1 was the day
of the BMT (range = 4 days prior to 10 days post), and 80% were completed by
posttransplantation day 3. Time 1 included measures of general distress, fear
network, perceived supportive and unsupportive responses. The second
assessment (Time 2) was conducted 3 months after the BMT/SCT and included the
same measures as Time 1. The mean date for Time 2 was 90 days post-BMT/SCT
(range = on time-21 days late), and 85% were completed within 10 days of the
3-month time point. The third assessment (Time 3) was conducted approximately
6 months after the BMT/SCT. Eighty percent of these assessments were completed
within 10 days of the 6-month time point (range = on time-1 month late). Time
3 included the same measures as Time 1, with the addition of a
paper-and-pencil PTSD symptoms measure and a PTSD structured interview. The
structured interview was administered to a subset of mothers. This strategy
was adopted because the time necessary to administer a structured interview
would add burden.
Two criteria were used to select the subset of mothers to administer the
psychiatric interview (SCID-NP-PTSD;
Spitzer et al., 1990). First,
a cutoff of 35 on the paper-and-pencil PTSD measure, the PCL-C, was used. This
cutoff was selected because cutoff scores between 35 and 40 on the PCL-C have
been shown to result in a low rate of false negative diagnoses of PTSD in
prior studies (Andrykowski, Cordova, &
Studts, 1998; Manne et al.,
1998). The good diagnostic utility of the PCL-C in terms of
minimizing false negatives suggested that using a cutoff on the PCL-C
minimizes the risk of missing a PTSD diagnosis. Second, 20% of the sample was
randomly preselected at Time 3 for psychiatric interview. Using these
criteria, 44 mothers were identified for SCID-NP administration. Twenty-four
mothers were selected because they had scores over 35 on the PCL-C.
Twenty-eight mothers were selected because they met random selection criteria.
There was an overlap between mothers receiving the SCID-NP by PCL-C cutoff and
random selection criteria. Eight mothers met both criteria. Four mothers who
met selection criteria did not complete the SCID interview either due to
refusal to be audiotaped or experimentor error. In total, 40 mothers were
interviewed with the SCID-NP.
Objective medical risk variables were obtained by review of the medical chart at all time points.
Measures
Posttraumatic Stress Disorder. The SCID-NP-PTSD
(Spitzer et al., 1990) has
four symptom clusters, Criterion A (stressor criterion), Criterion B
(reexperiencing cluster), Criterion C (avoidance cluster), and Criterion D
(arousal cluster). Symptoms must be present for more than 1 month (Criterion
E) and cause clinically significant distress or impairment in social,
occupational, or other functioning (Criterion F). All questions were
specifically linked to the child's BMT. The entire SCID was given, even if
mothers did not meet Criterion A. Interviewers were trained in the
SCID-NP-PTSD using the SCID-NP manual, audiotaped sample interviews, and
practice interviews. Additional training and supervision was provided during
the course of data collection. To establish interrater reliability, a subset
of 12 audiotapes was independently reviewed and scored by a second rater, a
licensed psychologist with a specialization in the SCID-NP (JD). Ratings were
compared using kappa coefficients and percent agreement. Standard criteria for
acceptable kappa levels do not exist. Fleiss
(1981) characterizes kappas of
.40 to .60 as fair, .60 to .75 as good, and greater than .75 as excellent.
Percent agreement for Criteria A, B, C, and D, and for the PTSD diagnosis,
were 1.0, 1.0, .92, 1.0, and 1.0, respectively. Under Fleiss's criteria,
results indicated that kappas for Criteria A, B, C, and D and PTSD diagnosis
were excellent (
= 1.0).
Posttraumatic Stress Symptoms. The Posttraumatic Symptom Disorder
Checklist—Civilian Version (PCL-C;
Weathers, Huska, & Keane,
1991) has 17 items and three subscales that correspond to DSM-IV
PTSD Criteria B, C, and D, as well as a total symptoms score. Items were keyed
to the child's BMT. The PCL-C can be used to identify individuals who may
merit PTSD diagnosis in two ways. The cutoff method suggests a total score of
50 or more as meriting formal diagnosis. The symptom cluster method suggests
that individuals may merit PTSD diagnoses if they report having been at least
moderately bothered by one or more reexperiencing, three or more avoidance,
and two or more arousal symptoms.
Trauma Exposure/Severity. Trauma exposure and severity were
assessed by measuring the mother's exposure to aversive medical experiences
during the child's BMT (labeled BMT events) and by assessing the medical risk
associated with the child's BMT (labeled BMT risk). The BMT events variable
was calculated at each time point. Time 1 events included the number of
infections during hospitalization, whether or not the child was transferred to
the ICU, placed on a ventilator, diagnosed with acute GVHD, or diagnosed with
cardiac, bladder, renal, pulmonary, hepatic, CNS, stomatitis, or
gastrointenstinal toxicities using the Bearman Common Toxicity Rating Scale
(Bearman et al., 1988). The 3-
and 6-month BMT events variable included the number of hospitalizations and
infections since the last assessment and whether the child was diagnosed with
chronic GVHD, or the toxicities listed above. The BMT/SCT risk variable was
formed by categorizing the BMT/SCT into three risk categories representing
increasing levels of risk: autologous BMT/SCT, allogeneic BMT/SCT with a
perfect HLA match, allogeneic transplant with an imperfect HLA match. This
variable was only calculated once.
Distress. The Beck Anxiety Inventory (BAI;
Beck, Epstein, Brown, & Steer,
1988) and the Beck Depression Inventory (BDI;
Beck, Steer, & Garbin,
1988) were used to assess distress at all time points. The BAI is
a 21-item self-report instrument used to assess symptoms of anxiety. The BAI
focuses on symptoms that are distinct from depressive symptoms
(Beck et al., 1988). In this
study, internal consistency was excellent (coefficient 
= .89, .87, .92
for Times 1, 2, and 3, respectively). The BDI is a widely used 21-item scale
used to assess depressive symptomatology. Items represent the presence and
intensity of emotional, cognitive, and somatic aspects of depression. In this
study, internal consistency was excellent (coefficient = .86, .87, and
.92 at Times 1, 2, and 3, respectively). Scores on both measures can range
from 0 to 63. For purposes of the regression analyses, BAI and BDI scores were
summed to create a single indicator of distress. This step was taken for two
reasons: First, we did not have separate predictions for anxiety and
depression. Second, collinearity diagnostics indicated that multicollinearity
was an issue, and a combined scale reduced this risk. The internal consistency
for the combined scale was excellent (coefficient = .91, .94, and .95
at Times 1, 2, and 3, respectively).
Fear Network. Based on prior research
(Foa & Kozak, 1986;
Lang, 1985), a measure of fear
network was developed. One item assessed the mother's perception of life
threat ("How scared are you that your child's treatment will not be
successful?") on a 9-point Likert scale. One item assessed the
perception of potential for suffering ("How scared are you that you'll
never be able to put the cancer experience behind you?") on a 9-point
Likert scale. Magnitude of fear was a sum of the intensity and frequency of
the mothers' fears. The intensity of fear was assessed by the number of
worries she had for her child in the domains of physical health, mental
health, social interactions, school activities, family interactions, and
future concerns. She was then asked to rate how frequently she had the worry
on a 7-point Likert scale. Fear network was a sum of life threat, potential
for suffering, number of fears, and the frequency of worries. Coefficient
alpha was satisfactory (Time 1 = .67, Time 2 = .68, Time 3 = .66).
Enacted Support. Thirteen items adapted from the Cancer Support
Inventory (Manne & Schnoll,
2001) were used. Likert-scale rated items assess emotional
support, information guidance, and assistance with adaptive coping. Mothers
were asked to make two separate ratings, one for their significant other and
one for other family and friends. Scores on each scale can range from 1 to 52.
The measure was administered at all time points. The internal consistency for
both ratings was excellent (partner, Time 1 = .85, Time 2 = .87, Time 3 = .91;
friends/family, Time 1 = .89, Time 2 = .87, Time 3 = .87).
Perceived Negative Behaviors. A 19-item scale adapted from the Cancer Support Inventory (Manne & Scholl, 2001) and Lepore's Social Constraints Measure (1996) was used. Items assessed overtly critical responses and more subtle avoidance. Mothers were asked to make separate ratings for their significant other and for other family and friends. Potential scores range from 1 to 76. The measure was administered at all time points. The internal consistency was excellent (partner, Time 1 = .91, Time 2 = .93, Time 3 = .95; friends/family, Time 1 = .91, Time 2 = .88, Time 3 = .93).
Overview of Data Analysis
Study analyses were conducted in three steps. The first step was to provide
descriptive information for model predictors and outcomes and to evaluate
changes in model variables over time. Changes in model variables were
evaluated using repeated measures analysis of variance. The second step was to
identify a set of variables for inclusion in the multivariable analyses
predicting PTSD. This step was taken to reduce the potential number of
variables (n = 16) entered into the multivariable analyses. Bivariate
analyses evaluating the association between the outcome and demographic,
medical, and psychological predictor variables were conducted. Demographic
variables included child age, child gender, maternal age, education (completed
high school vs. did not complete high school), marital status (married vs. not
married), ethnicity (Caucasian vs. non-Caucasian), and family income. Medical
variables included hospital recruited from (six categories), type of illness
(cancer vs. non-malignant condition), HLA compatability status (perfect match
vs. mismatch in one to three antigens), and time from diagnosis to BMT.
Psychological variables included distress, enacted support, and negative
responses. For the PCL-C symptom total, t tests or analysis of
variance were conducted for categorical descriptor variables, and correlations
were conducted for continuous variables. For the SCID-PTSD diagnosis,
chi-square was used for the categorical descriptor variables and t
tests were conducted for continuous variables.
Once a set of predictors was identified, the third step was to use multivariable analyses to evaluate the relative contribution of model variables in predicting PTSD diagnosis and symptoms. For this analysis, separate hierarchical regression analyses were carried out with 6-month PCL-C total scores and SCID-NP-PTSD diagnosis as outcomes. Analyses were conducted separately with model predictor variables from Times 1, 2, and 3. Trauma exposure/severity was entered into every regression equation to control for this factor. Variables were entered into the equations in the following order: trauma exposure/severity, psychological distress, fear network, and social processing variables. Social processing variables were entered into the equation after cognitive appraisals because we assumed that contextual responses occurred temporally after cognitive appraisals.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Descriptive Data and Changes Over 6 Months
Based on SCID-NP-PTSD responses, seven mothers (7.8 % of all participants and 17.5% of the 40 participants completing the PTSD module) were diagnosed with current PTSD. Partial PTSD, which is defined as meeting the stressor Criterion A and criteria for two of the three PTSD symptom clusters, was present in an additional six mothers (6.6% of all participants and 15% of the 40 mothers completing the PTSD module). Among the 40 mothers administered the SCID-NP, Criterion A was met by 23 mothers (57.5% of mothers completing PTSD module), Criterion B was met by 17 mothers (39.3% of mothers completing the PTSD module), Criterion C was met by 15 mothers (37% of mothers completing the PTSD module), and Criterion D was met by 8 mothers (29.6% of mothers completing the PTSD module). The most frequently endorsed symptoms were "difficulty concentrating" (52%), "difficulty sleeping" (60%), "feelings of detachment" (48%), "irritability or angry outbursts" (44%), "recurrent or distressing recollections of the experience" (43.3%), and "a sense of foreshortened future" (38%). The least endorsed symptoms were "inability to recall important aspects of the trauma" (14.8%), "restricted range of affect" (18.5%), and "acting or feeling as if the event were recurring" (20%). Among the six mothers who met partial PTSD criteria, five mothers met criteria for Criterion B and six mothers met criteria for Criterion D. Only one mother met criteria for Criterion C (avoidance and numbing).
Mothers reported a mean PCL-C score of 27.5 with a standard deviation of 10.08. Using a score of 50 or greater total score on the PCL-C, we found that only three mothers met criteria for PTSD (3.3%). Using the symptom cluster method of determining PTSD, four mothers met criteria. One mother met symptom cluster criteria but did not have a PCLC total score of greater than 50.
Descriptive information for the model predictor variables as well as results of the repeated measures analyses of variance are shown in Table I. BAI and BDI scores are presented as well as the combined distress score. Results of repeated measures analyses of variance revealed that the BDI, BAI, and overall distress evidenced significant reductions over time. The number of BMT events also evidenced a significant reduction over time, as did the amount of perceived family criticism. Enacted support and fear network did not evidence changes over time.
|
Analysis of Predictors of PTSD
Predictors of PCL-C Symptom Total. Bivariate analyses did not
reveal significant associations between demographic or medical variables and
PCL-C total scores. Thus, other than the trauma exposure/severity variable,
these variables were not included in the multivariable regressions. Bivariate
correlations between model predictors at all time points and 6-month PCL-C
scores are shown in Table II.
The bivariate correlations revealed that Time 1, 2, and 3 fear network,
psychological distress, and partner and family and friend negative responses
had significant correlations with PCL-C total symptoms. Partner and family and
friend support did not evidence significant correlations with PCL-C total and
were not entered in the multivariable regression equations.
|
The results of the regression analyses are shown in Table III. For the equation using Time 1 predictors, fear network, distress, and negative responses from friends and family predicted PTSD symptoms. The trauma severity variables were not a significant predictor, accounting for only 1.2% of the variance in PTSD symptoms. Time 1 distress accounted for an additional 14.6% of the variance in PTSD symptoms. Fear network accounted for an additional 25.1% of the variance in PTSD symptoms, and the block that included family criticism (along with partner criticism) predicted an additional 7.5% of the variance.
|
The set of predictors accounted for 48.4% of the variance in PCL-C scores.
In the equation using Time 2 predictors, only distress accounted for a significant amount of variance in PTSD symptoms (37.8%). In the equation using concurrent predictors (Time 3), psychological distress was a significant correlate of PTSD symptoms, accounting for nearly 61% of the variance in PCLC total score. Trauma severity and critical responses from partner or family members were not associated with PCLC total (although family negative responses evidenced a marginal association), once the other variables were entered into the equation. The set of predictors accounted for slightly more than 64% of the variance in PCL-C scores.
Predictors of SCID-NP-PTSD Diagnosis. For these analyses, we used a dichotomous variable indicating whether the mother received a diagnosis of PTSD. Bivariate analyses using series of chi-squares and t tests indicated that none of the demographic or medical factors was significantly associated with PTSD diagnosis. Among the Time 1 variables, baseline distress (t[38] = 2.65, p < .05), fear network (t[38] = 4.31, p < .001), partner negative responses (t [38] = 2.15, p < .05), and friends and family negative responses (t [38] = 2.1, p < .05) were associated with PTSD. Among the Time 2 variables, distress (t [38] = 3.2, p < .01) and partner negative responses (t [38] = 2.7, p < .01) were significantly associated with PTSD. Among the Time 3 variables, only psychological distress was significantly associated with PTSD (t [38] = 4.0, p < .001). Based on these analyses, hierarchical logistic regressions were conducted using each set of predictors.
The results of these analyses are shown in Table IV. None of the Time 1 or 2 predictors was significantly associated with PTSD diagnosis. Time 3 distress remained a significant predictor of PTSD diagnosis after accounting for trauma severity variables.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
The results support the role of cognitive and social processing factors assessed at transplantation in PTSD symptoms among mothers of children under-going BMT. Fear structure, distress, and unsupportive responses by family and friends measured at transplantation were predictive of PTSD symptom severity 6 months after BMT. These data underscore the potential for predicting PTSD symptom severity from maternal fear appraisals, distress, and perceived negative responses from others during the child's BMT. However, these variables were less successful in predicting PTSD symptoms when measured at later time points and in predicting a formal PTSD diagnosis as assessed by the SCID-NP. Only anxiety and depressive symptoms were predictive of PTSD symptom severity, and concurrent psychological distress was the only variable associated with a PTSD diagnosis at 6 months.
Cognitive processing at the time of transplantation played the most
important role in later PTSD symptoms. Mother's appraisal of threat to her
child's life, potential for her personal suffering, and the number of fears
she reported regarding her child's future functioning at the time of the
transplantation were the strongest predictors of PTSD symptoms. These results
are consistent with our hypotheses and with prior research conducted by Kazak
et al. (1998). Emotional
distress at the time of BMT was also a key predictor of PTSD symptoms. These
results are consistent with research with other populations indicating that
distress during or immediately after a traumatic event are associated with
PTSD symptoms (Maercker et al.,
2000; Roemer, Orsillo,
Borkovec, & Litz, 1998).
Family and friend negative responses at BMT were predictive of later PTSD
symptoms. The role of negative responses in PTSD has received relatively
little attention in the PTSD or pediatric BMT and cancer literature. Our
cross-sectional study of mothers of pediatric cancer survivors indicated that
social constraints were associated with PTSD symptoms
(Manne et al., 2000). Dunmore,
Clark, and Ehlers (1999)
studied assault victims and found that participants with PTSD were more likely
to perceive that others responded negatively to them than participants who did
not have PTSD. Maternal perceived family and friend criticism may reflect
general overall family functioning. If so, then mothers who had problems
managing the stress of the actual BMT may be at greater risk for PTSD
symptoms.
Despite the association between cognitive and social processing variables assessed at the time of BMT and later PTSD symptoms, these variables were less predictive when measured 3 and 6 months after BMT. Psychological distress assessed 3 and 6 months post-BMT was associated with PTSD symptoms, but fear network and negative responses from friends and family assessed at these time points were not associated with PTSD symptoms. Moreover, cognitive and social processing variables other than concurrent emotional distress were not able to predict a formal PTSD diagnosis. The lack of prospective associations may indicate that appraisals and reactions at the time of the actual transplantation may be most important in the development of later PTSD symptoms. The findings may also be due to the small sample size of mothers with PTSD and the number of other variables placed into the regression equations in the prospective analyses, and a larger sample of mothers with PTSD may have yielded different findings. However, our findings also raise the possibility that the nature and distribution of PTSD may differ among individuals who witness a traumatic event and suggest future studies should carefully evaluate differences in the nature of PTSD among witnesses to trauma.
Two variables, social support and exposure to BMT events and medical risk,
were not associated with PTSD symptoms. This finding is not consistent with
other studies (e.g., Joseph, Andrews,
Williams, & Yule, 1992;
Manne, DuHamel, & Redd,
2000). One explanation may be that prior research evaluated
perceived social support. Studies that have evaluated the role of received
social support have reported inconsistent findings, with some studies not
finding an association with PTSD (Pickens,
Field, Prodromidis, Nogueras, & Hossain, 1995) and some
findings reporting a relation between greater received support and lower PTSD
(Stretch, 1986). These
findings suggest that greater attention should be given to the role of various
types of social support, as different dimensions may have different
associations with PTSD. The fact that we did not find an association between
exposure to aversive BMT events and medical risk of the child's BMT and PTSD
symptoms suggests that subjective appraisal may be more important than
objective factors. This finding is consistent with research by Creamer et al.
(1992) that failed to find a
link between exposure and other "severity" factors and PTSD.
However, other studies have found an "exposure-response"
relationship (e.g., Fontana &
Rosenheck, 1993). In addition, other studies of parents of
childhood cancer survivors have documented a link between the length of cancer
treatment and PTSD (Kazak et al.,
1998). There are at least two alternative explanations for our
negative finding. First, our measure of stressor severity and exposure may not
have been sufficiently comprehensive to capture the objective aspects of BMT
that result in PTSD. Second, the link between exposure/severity and PTSD may
be mediated by other mechanisms such as appraisal of life threat or treatment
intensity (Kazak et al.,
1998).
This study has several limitations. First, present PTSD symptoms could reflect preexisting symptoms present prior to the transplantation. A second, related issue is that present PTSD symptoms were actually cued by events prior to the transplantation. Future studies should evaluate symptoms in the several months prior to the child's transplantation. However, even if symptoms prior to the transplantation were assessed, to some degree the post-BMT assessment would be confounded by pre-BMT experiences. PTSD symptoms can reflect a multiplicity of different cues and stressors, and it is difficult for the participant, as well as the SCID interviewer, to separate cues for PTSD symptoms. Third, mothers who did not complete the 6-month survey may have been more likely to report PTSD symptoms. Mothers with PTSD symptoms may avoid completing a survey and interview that would serve as a reminder of the BMT/SCT. Comparisons of mothers who did not complete the second or third assessments were consistent with this hypothesis: mothers who dropped out, for reasons other than the child's death after the BMT/SCT, were more likely to report anxiety and depressive symptoms and a higher fear network score at Time 1. Attrition may affect estimates of prevalence and severity of PTSD symptoms (e.g., higher participation might result in more mothers with PTSD at 6 months). Biased attrition may also affect relations between our model variables and PTSD. Overall, our findings suggest that mothers at highest risk for developing PTSD may not participate in psychosocial studies. Both clinicians and researchers should be aware that we may know less about those mothers who are most at risk for developing PTSD.
In this study, we identified factors that place mothers at risk for developing PTSD symptoms, but we were less able to identify factors that lead to a formal diagnosis of PTSD. Mothers at risk for later PTSD symptoms have more distress at the time of BMT/SCT, report more negative responses from friends and family, and perceive more life threat to their child, more potential for their own personal suffering, and more fears about the child's future functioning. Our findings also suggest that mothers who drop out of our research might be at higher risk for developing PTSD. This finding suggests that interventions delivered early in transplantation might be able to target and reduce distress among a group of mothers who are less likely to participate in psychosocial studies or interventions later.
|
Acknowledgments |
|---|
This work was supported by grant MH57738 from the National Institutes of Health.
Received May 30, 2001; revision received October 10, 2001; revision received December 4, 2001; revision received February 1, 2002; accepted February 8, 2002
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Andrykowski, M. A., Cordova, M. J., & Studts, J. L. (1998). Diagnosis of posttraumatic stress disorder following treatment for breast cancer. Journal of Consulting and Clinical Psychology, 66, 586-590.[Web of Science][Medline]
Balduzzi, A., Gooley, T., Anasetti, C., Sanders, J. E., Martin, P. J., & Petersdorf, E. W. (1994). Unrelated donor marrow transplantation in children. Blood, 61(Suppl. 1), 393a.
Barrera, M., Boyd Pringle, L., Sumbler, K., & Saunders, F. (2000). Quality of life and behavioral adjustment after pediatric bone marrow transplantation. Bone Marrow Transplantation, 26, 427-435.[Web of Science][Medline]
Bearman, S., Appelbaum, F., Buckner, C., Petersen, F., Fisher, L.,
Clift, R., & Thomas, E. D. (1988). Regimen-related toxicity
in patients undergoing bone marrow transplantation. Journal of
Clinical Oncology, 6(10),
1562-1568.
Beck, A. T., Epstein, N., Brown, G., & Steer, R. A. (1988). An inventory for measuring clinical anxiety: Psychometric properties. Journal of Clinical and Consulting Psychology, 56, 893-897.
Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8, 77-100.
Butler, R., Rizzi, L., & Handwerger, B. (1996).
Brief report: The assessment of posttraumatic stress disorder in pediatric
cancer patients and survivors. Journal of Pediatric
Psychology, 21,
499-504.
Cardena, E., Koopman, C., Classen, C., Waelde, L. D., & Spiegel, D. (2000). Psychometric properties of the Stanford acute stress reaction question. Journal of Traumatic Stress, 13(4), 719-734.[Web of Science][Medline]
Creamer, M., Burgess, P., & Pattison, P. (1992). Reaction to trauma: A cognitive processing model. Journal of Abnormal Psychology, 101, 452-459.[Web of Science][Medline]
Dunmore, E., Clark, D., & Ehlers, A. (1999). Cognitive factors involved in the onset and maintenance of posttraumatic stress disorder (PTSD) after physical or sexual assault. Behavior Research and Therapy, 37, 809-829.[Web of Science][Medline]
Fleiss, J. (1981). Statistical methods for rates and proportions. New York: Wiley.
Foa, E., & Kozak, M. (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 99, 20-35.[Web of Science][Medline]
Fontana, A., & Rosenheck, R. (1993). A causal model of the etiology of war-related PTSD. Journal of Traumatic Stress, 6, 475-500.
Heiney, S. P., Neuberg, R. W., Myers, D., & Bergman, L. H. (1994). The aftermath of bone marrow transplant for parents of pediatric patients: A post-traumatic stress disorder. Oncology Nursing Forum, 21(5), 843-847.[Medline]
Joseph, S. A., Andrews, B., Williams, R., & Yule, W. (1992). Crisis support and psychiatric symptomatology in adult survivors of the Jupiter cruise ship disaster. British Journal of Clinical Psychology, 31, 63-74.
Kazak, A., Barakat, L., Meeske, K., Christakis, D., Meadows, A., & Casey, R. (1997). Posttraumatic stress, family functioning, and social support in survivors of childhood leukemia and their mothers and fathers. Journal of Consulting and Clinical Psychology, 65, 120-129.[Web of Science][Medline]
Kazak, A., Stuber, M., Barakat, L., Meeske, K., Guthrie, D., & Meadows, A. (1998). Predicting posttraumatic stress symptoms in mothers and fathers of survivors of childhood cancers. Journal of American Academy of Child and Adolescent Psychiatry, 37, 823-831.[Web of Science][Medline]
Kliewer, W., Lepore, S., Oskin, D., & Johnson, P. (1998). The role of social and cognitive processes in children's adjustment to community violence. Journal of Consulting and Clinical Psychology, 66, 220-225.
Lang, P. J. (1985). The cognitive psychophysiology of emotion: Fear and anxiety. In A. H. Tuma & J. D. Maser (Eds.), Anxiety and the anxiety disorders (pp. 131-170). Hillsdale, NJ: Erlbaum.
Leisner, R. J., Leiper, A. D., Hann, I., & Chessels, P.
(1994). Late effects of intensive treatment for acute myeloid
leukemia and myelodysplasia in childhood. Journal of Clinical
Oncology, 12(5),
916-924.
Lepore, S. J., Silver, R. C., Wortman, C. B., & Wayment, H. (1996). Social constraints, intrusive thoughts, and depressive symptoms among bereaved mothers. Journal of Personality and Social Psychology, 70, 271-282.[Web of Science][Medline]
Maercker, A., Beauducel, A., & Schutzwohl, M. (2000). Trauma severity and initial reactions as precipitating factors for posttraumatic stress symptoms and chronic dissociation in former political prisoners. Journal of Traumatic Stress, 13(4), 651-660.[Web of Science][Medline]
Manne, S., DuHamel, K., Gallelli, K., Sorgen, K., & Redd, W. H.
(1998). Post-traumatic stress disorder among mothers of pediatric
cancer survivors: Diagnosis, comorbidity, and utility of the PTSD checklist as
a screening instrument. Journal of Pediatric Psychology,
23(6),
357-365.
Manne, S., DuHamel, K., & Redd, W. H. (2000). Association of psychological vulnerability factors to post-traumatic stress symptomatology in mothers of pediatric cancer survivors. Psycho-Oncology, 9, 372-384.[Medline]
Manne, S., & Schnoll, R. (2001). Measuring supportive and unsupportive responses during cancer treatment: A factor analysis of the partner responses to cancer inventory. Journal of Behavioral Medicine, 24, 297-322.[Web of Science][Medline]
March, J. S. (1993). What constitutes a stressor? The criterion A issue. In J. Davidson & E. Foa (Eds), Posttraumatic stress disorder: DSM-IV and beyond (pp. 37-54). Washington, DC: American Psychiatric Press.
Pelcovitz, D., Goldenberg, B., Kaplan, S., Weinblatt, M., Mandel,
F., Meyers B., & Vinciguerra, V. (1996). Posttraumatic stress
disorder in mothers of pediatric cancer survivors.
Psychosomatics, 37,
116-126.
Pickens, J., Field, T., Prodromidis, M., Nogueras, M., & Hossain, Z. (1995). Postraumatic stress, depression and social support among college students after Hurricane Andrew. Journal of College Student Development, 36(2), 152-161.
Roemer, L., Orsillo, S., Borkovec, T., & Litz, B. (1998). Emotional response at the time of a potentially traumatizing event and PTSD symptomatology: A preliminary retrospective analysis of the DSM-IV Criterion A-2. Journal of Behavior Therapy, 29, 123-130.
Spitzer, R. L., Williams, J. W., Gibbon, M., & First, M. B. (1990). Structured clinical interview for DSM-III-R-Non-Patient Edition (SCID-NP Version 1.0). Washington, DC: American Psychiatric Press.
Streisand, R., Rodrigue, J., Houck, C., Graham-Pole, J., &
Berlant, N. (2000). Brief report: Parents of children undergoing
bone marrow transplantation: Documenting stress and piloting a psychological
intervention program. Journal of Pediatric Psychology,
25, 331-337.
Stretch, R. H. (1986). Incidence and etiology of post-traumatic stress disorder among active duty army personnel. Journal of Applied Social Psychology, 16(6), 464-481.
Tait, R., & Silver, R. (1992). Coming to terms with major negative life events. In J. S. Uleman & J. Bargh (Eds.), Unintended thought (pp. 351-382). New York: Guilford Press.
Weathers, F. W., Huska, J. A., & Keane, T. M. (1991). The PTSD checklist-civilian version (PCL-C). (Scale available from Sharon Manne at the National Center for PTSD, Boston VA Medical Center, Boston, MA.)
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Manne Commentary: Adopting to a Broad Perspective on Posttraumatic Stress Disorders, Childhood Medical Illness and Injury J. Pediatr. Psychol., January 1, 2009; 34(1): 22 - 26. [Full Text] [PDF]
|
|
|
|
|
|
A. E. Kazak, M. T. Rourke, M. A. Alderfer, A. Pai, A. F. Reilly, and A. T. Meadows Evidence-based Assessment, Intervention and Psychosocial Care in Pediatric Oncology: A Blueprint for Comprehensive Services Across Treatment J. Pediatr. Psychol., October 1, 2007; 32(9): 1099 - 1110. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Stoppelbein, L. Greening, and T. D. Elkin Risk of Posttraumatic Stress Symptoms: A Comparison of Child Survivors of Pediatric Cancer and Parental Bereavement J. Pediatr. Psychol., May 1, 2006; 31(4): 367 - 376. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Kazak, N. Kassam-Adams, S. Schneider, N. Zelikovsky, M. A. Alderfer, and M. Rourke An Integrative Model of Pediatric Medical Traumatic Stress J. Pediatr. Psychol., May 1, 2006; 31(4): 343 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Phipps, S. Larson, A. Long, and S. N. Rai Adaptive Style and Symptoms of Posttraumatic Stress in Children with Cancer and Their Parents J. Pediatr. Psychol., April 1, 2006; 31(3): 298 - 309. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Kazak, S. Simms, M. A. Alderfer, M. T. Rourke, T. Crump, K. McClure, P. Jones, A. Rodriguez, A. Boeving, W.-T. Hwang, et al. Feasibility and Preliminary Outcomes from a Pilot Study of a Brief Psychological Intervention for Families of Children Newly Diagnosed with Cancer J. Pediatr. Psychol., December 1, 2005; 30(8): 644 - 655. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Kazak, C. A. Boeving, M. A. Alderfer, W.-T. Hwang, and A. Reilly Posttraumatic Stress Symptoms During Treatment in Parents of Children With Cancer J. Clin. Oncol., October 20, 2005; 23(30): 7405 - 7410. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. L. Mintzer, M. L. Stuber, D. Seacord, M. Castaneda, V. Mesrkhani, and D. Glover Traumatic Stress Symptoms in Adolescent Organ Transplant Recipients Pediatrics, June 1, 2005; 115(6): 1640 - 1644. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Phipps, M. Dunavant, S. Lensing, and S. N. Rai Psychosocial Predictors of Distress in Parents of Children Undergoing Stem Cell or Bone Marrow Transplantation J. Pediatr. Psychol., March 1, 2005; 30(2): 139 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Kazak Evidence-Based Interventions for Survivors of Childhood Cancer and Their Families J. Pediatr. Psychol., January 1, 2005; 30(1): 29 - 39. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Manne, K. DuHamel, J. Ostroff, S. Parsons, D. R. Martini, S. E. Williams, L. Mee, S. Sexson, J. Austin, J. Difede, et al. Anxiety, Depressive, and Posttraumatic Stress Disorders Among Mothers of Pediatric Survivors of Hematopoietic Stem Cell Transplantation Pediatrics, June 1, 2004; 113(6): 1700 - 1708. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Kazak, M. Alderfer, M. T. Rourke, S. Simms, R. Streisand, and J. R. Grossman Posttraumatic Stress Disorder (PTSD) and Posttraumatic Stress Symptoms (PTSS) in Families of Adolescent Childhood Cancer Survivors J. Pediatr. Psychol., April 1, 2004; 29(3): 211 - 219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. N. DuHamel, S. Manne, N. Nereo, J. Ostroff, R. Martini, S. Parsons, S. Williams, L. Mee, S. Sexson, L. Wu, et al. Cognitive Processing Among Mothers of Children Undergoing Bone Marrow/Stem Cell Transplantation Psychosom Med, January 1, 2004; 66(1): 92 - 103. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||